The role of the tau protein in neurological diseases is an important open question, given that it has been demonstrated as a central driver of neurodegeneration. Though tau's pathogenic mechanisms are still unclear, variants in the MAPT gene, encoding for tau, have been identified as causal in some tauopathies, and the H1 haplotype at the 17q21.31 locus (containing MAPT) has been strongly linked to risk for progressive supranuclear palsy (PSP), which is pathologically the purest tauopathy. However, the use of traditional techniques such as genome-wide association studies (GWAS) has stalled in advancing this field due to limited scope of data and lack of statistical power. Meanwhile, emerging technologies such as next-generation sequencing allow the generation of richer sets of genetic, epigenetic, and gene expression data. Using such methods, we have discovered potential factors on the molecular level that contribute to certain tauopathies in human blood cells. In this proposal we aim to expand and leverage these diverse datasets in an integrative fashion to overcome the limitations of traditional statistical approaches and provide mechanistic insight into the pathogenicity of tau protein. First, we will extend our analysis of H1 haplotype differential gene expression and epigenetic changes from blood cells to human brain regions, using RNA-seq of cells from multiple brain regions of normal subjects. These findings will be confirmed and validated for involvement in disease using real-time PCR experiments in brains from PSP cases. Second, we will detect novel alleles outside of the tau locus that modify risk for PSP. Besides the fact that our cohort is the only dataset of PSP patients for which whole-genome sequencing is available, we additionally have data regarding DNA methylation and gene expression from the same subjects. Integrative analyses of such data using methods of rare variant association, clustering, and causality analysis will be able to discover effects which were previously undetectable, and provide biological and functional context. Third, we will use genomic editing by transcription activator-like effector nucleases (TALENs) to introduce candidate mutations into cell lines. This experiment will thereby generate cell models of tauopathy, experimentally validate mechanistic findings, and provide strong causal evidence. By performing this work, we will discover genetic factors that contribute to tauopathies and the mechanisms by which they operate, findings of diagnostic and therapeutic significance.